This text is based on a text originally written in Dutch in 2005 for
the
benefit of people who don't have much familiarity with my work but who
had to judge my scores. I think however that it might be interesting
for everybody who'd like to know something about how I work and think.
So here's an English version, for the entire Internet to enjoy.
Texture, melody, meter
In composing, I have often aimed at a kind of panoramic polyphony. From
my piece Toccata III for 2 Glockenspiels (2001) on especially, my
technical ideas on this developped gradually. By "panorama" I mean: a
complete equality of all voices. A number of voices are moving
simultaneously through some musical space, and in the process, no
single voice is carrying the musical argument more than any other one,
none is functioning as a central voice.
In my panoramas I usually try to achieve a certain homogeneity or
compactness. Every time I apply this principle, what I try to achieve
is the possibility that the voices become entangled for the ear, so
that different voices may seem to feed into one another. Clear
differentations of function, like melody and accompaniment, are
avoided. The voices in the panorama move within the same parametric
space, or within closely related parametric spaces: it should be
possible to relate the rhythm, articulation, timbre, tempo and dynamics
of different voices to each other. The melodic material of the voices,
as well, is preferably strongly related or even identical, so that you
can find mirrorings in the melodic realm as well (canons, imitations,
heterophony.)
Melodically, it often turns out to be necessary to write extremely
simple, almost primitive shapes. Within the melodic material itself I
often opt for possibilities to lead the ear along different paths at
the same time using simple means. Among these a conscious and
consistent use of Sekundgang (relations between consecutive top and
bottom notes in compound melodic contours.) Melodies are often put
together as mosaics of small, atomic note-groups, in which every
note-group is characterized by a very simple and therefore very clear
type of motion (ascending, descending, alternating between two notes,
etc.)
My melodic structure often contains a metrical character, which
corresponds to the melodic contour. A metrical unit will coincide with
an
atomic group of notes with a characteristic
motion-type, and whenever the melody goes from one such grouplet to the
next
this is heard as a metrical time-point. Even if I hardly ever write
time
signatures and even if in many spots, I have used exclusively eighth
notes in my notation, I'm always conscious of this metrical character
and of rhythm. A meter can arise from a simple repetition of melodic
motives. If, within one panorama, different voices, which all have
their own metrical character or their own tempo, enter into an audible
relationship to one another, what I call a meta-meter may result: a complex
metrical experience in which individual metrical
layers are submerged.
For listeners, all this leads to a free, multi-dimensional sonic
experience. In many pieces of mine listeners are given the possibility
either to follow specific voices, or to follow relationships between
certain voices, or to follow the resultant total: but it's hardly
possible to follow all aspects at the same time.
Form
In using this type of texture, which is all about an internal motion
that happens between a number of voices, it's often less desirable to
use a type of form that depends primarily on well-timed rhetorical
gestures and impressive dramatic shifts. Rather, I would give the
listener some time so that he or she can find his or her way within the
panoramic texture. In the most extreme cases, form is entirely an
emerging phenomenon that happens in the listener's mind: emerging from
the way his or her attention is ambling through musical space, all by
itself, just by listening.
Coordination
between parts
In many pieces of recent years I have dropped the idea of written-out
coordination between parts. I feel that the kind of counting that
goes with standard notation is not what my music is about. It is not to
my taste to see a group of musicians get the coordination of their
parts right by moving together according to a meter that you don't hear
and which the composer doesn't even want you to hear, and I certainly
dislike the related phenomenon, all too common in new music, of the
conductor as a luxury type of metronome. Fixations on the wrong kind of
rhythmical precision can in fact also be bad for the beauty of the
sound itself.
For that reason, over the years I have explored ways to coordinate
parts differently, for example through the use of stopwatch, or of cues
etc. in a great diversity of variations. Of course, with techniques
like these in which minute details of coordination are no longer fixed
a score is no longer necessary. Generally, in works in which there's no
role for a conductor I no longer write a score - there are only parts.
Notation
Notation is not meant to describe a sound, but to instruct a performer.
I make many notational decisions based on what kind of playing they
suggest - my method of instruction will influence the attitude of the
performer, and therefore it will influence the character, the
atmosphere, of a performance. Anything that I do not need towards this
goal I will try not to notate. For example, I'll only write a time
signature if I want there to be explicit metrical counting independent
of melodic shape (as indicated above, in many cases meter is already an
implicit part of the melodic shape.) Whenever it is possible to get to
a musical result by using a suggestive verbal explanation instead of a
fully fleshed-out notation I prefer the verbal instruction.
Often, I try to maintain a very basic 'notational vocabulary' in a
piece
or in a movement. The decision as to which notational means I'm going
to employ and which I'm not going to employ is to me part of the
composition. Whatever I choose not to notate is as much part of
compositional decision-making as are the things I do choose to notate.
Indeterminacy
In certain pieces written over the past few years certain aspects of
performance are left indeterminate. Whenever I do this, I make sure
that performers know exactly what decision he or she is supposed to
take, and what kind of difference such performer decisions will make.
In such pieces, I do not compose specific sounds, but I do compose the
decision process of the performers and by that, a psychology or
theatrical charge. A complex decision that has to be made in little
time for instance may be an instrument of tension.
The choices that a performer has to make are never arbitrary. It's
never just a choice between, say, three notes that could just as well
have been three other notes, but it's always a choice with a dramatic
meaning. Often it involves timing, and often in such a way that
performers will have to react to one another (this happens in the
middle part of Eindig Stuk for example.)
Refined excess
The intention with all of this is to reach, in a controlled way, an
unknowably rich musical experience. A subtle instability of the musical
event. To lead a musical language that is in principle clear and
accessible to the limits of its own perceptibility. To cause excesses
of meaningful, comprehensible motions and nuances from simple
materials. A music that is perfectly logical in ways that are not
entirely controllable. A friendly confusion. A refined excess.
For more on metametrics, Kyle Gann's weblog has many
entries on the subject - in fact his weblog is the place where the word
metametrics was coincd